Utilise clock provider API with struct clk_hw instances instead of the
consumer-side struct clk.
Signed-off-by: Harry Austen <hpausten@protonmail.com>
---
drivers/clk/xilinx/clk-xlnx-clock-wizard.c | 101 ++++++++-------------
1 file changed, 39 insertions(+), 62 deletions(-)
diff --git a/drivers/clk/xilinx/clk-xlnx-clock-wizard.c b/drivers/clk/xilinx/clk-xlnx-clock-wizard.c
index 0ca045849ea3e..30c5cc9dcd7e9 100644
--- a/drivers/clk/xilinx/clk-xlnx-clock-wizard.c
+++ b/drivers/clk/xilinx/clk-xlnx-clock-wizard.c
@@ -17,6 +17,7 @@
#include <linux/of.h>
#include <linux/math64.h>
#include <linux/module.h>
+#include <linux/overflow.h>
#include <linux/err.h>
#include <linux/iopoll.h>
@@ -121,24 +122,22 @@ enum clk_wzrd_int_clks {
/**
* struct clk_wzrd - Clock wizard private data structure
*
- * @clk_data: Clock data
+ * @clk_data: Output clock data
* @nb: Notifier block
* @base: Memory base
* @clk_in1: Handle to input clock 'clk_in1'
* @axi_clk: Handle to input clock 's_axi_aclk'
* @clks_internal: Internal clocks
- * @clkout: Output clocks
* @speed_grade: Speed grade of the device
* @suspended: Flag indicating power state of the device
*/
struct clk_wzrd {
- struct clk_onecell_data clk_data;
+ struct clk_hw_onecell_data *clk_data;
struct notifier_block nb;
void __iomem *base;
struct clk *clk_in1;
struct clk *axi_clk;
- struct clk *clks_internal[wzrd_clk_int_max];
- struct clk *clkout[WZRD_NUM_OUTPUTS];
+ struct clk_hw *clks_internal[wzrd_clk_int_max];
unsigned int speed_grade;
bool suspended;
};
@@ -765,7 +764,7 @@ static const struct clk_ops clk_wzrd_clk_divider_ops_f = {
.recalc_rate = clk_wzrd_recalc_ratef,
};
-static struct clk *clk_wzrd_register_divf(struct device *dev,
+static struct clk_hw *clk_wzrd_register_divf(struct device *dev,
const char *name,
const char *parent_name,
unsigned long flags,
@@ -805,10 +804,10 @@ static struct clk *clk_wzrd_register_divf(struct device *dev,
if (ret)
return ERR_PTR(ret);
- return hw->clk;
+ return hw;
}
-static struct clk *clk_wzrd_ver_register_divider(struct device *dev,
+static struct clk_hw *clk_wzrd_ver_register_divider(struct device *dev,
const char *name,
const char *parent_name,
unsigned long flags,
@@ -852,10 +851,10 @@ static struct clk *clk_wzrd_ver_register_divider(struct device *dev,
if (ret)
return ERR_PTR(ret);
- return hw->clk;
+ return hw;
}
-static struct clk *clk_wzrd_register_divider(struct device *dev,
+static struct clk_hw *clk_wzrd_register_divider(struct device *dev,
const char *name,
const char *parent_name,
unsigned long flags,
@@ -898,7 +897,7 @@ static struct clk *clk_wzrd_register_divider(struct device *dev,
if (ret)
return ERR_PTR(ret);
- return hw->clk;
+ return hw;
}
static int clk_wzrd_clk_notifier(struct notifier_block *nb, unsigned long event,
@@ -1020,13 +1019,18 @@ static int clk_wzrd_probe(struct platform_device *pdev)
if (ret || nr_outputs > WZRD_NUM_OUTPUTS)
return -EINVAL;
+ clk_wzrd->clk_data = devm_kzalloc(&pdev->dev, struct_size(clk_wzrd->clk_data, hws,
+ nr_outputs), GFP_KERNEL);
+ if (!clk_wzrd->clk_data)
+ return -ENOMEM;
+
clkout_name = devm_kasprintf(&pdev->dev, GFP_KERNEL, "%s_out0", dev_name(&pdev->dev));
if (!clkout_name)
return -ENOMEM;
if (is_versal) {
if (nr_outputs == 1) {
- clk_wzrd->clkout[0] = clk_wzrd_ver_register_divider
+ clk_wzrd->clk_data->hws[0] = clk_wzrd_ver_register_divider
(&pdev->dev, clkout_name,
__clk_get_name(clk_wzrd->clk_in1), 0,
clk_wzrd->base, WZRD_CLK_CFG_REG(is_versal, 3),
@@ -1059,7 +1063,7 @@ static int clk_wzrd_probe(struct platform_device *pdev)
div = 64;
} else {
if (nr_outputs == 1) {
- clk_wzrd->clkout[0] = clk_wzrd_register_divider
+ clk_wzrd->clk_data->hws[0] = clk_wzrd_register_divider
(&pdev->dev, clkout_name,
__clk_get_name(clk_wzrd->clk_in1), 0,
clk_wzrd->base, WZRD_CLK_CFG_REG(is_versal, 3),
@@ -1082,7 +1086,7 @@ static int clk_wzrd_probe(struct platform_device *pdev)
clk_name = devm_kasprintf(&pdev->dev, GFP_KERNEL, "%s_mul", dev_name(&pdev->dev));
if (!clk_name)
return -ENOMEM;
- clk_wzrd->clks_internal[wzrd_clk_mul] = clk_register_fixed_factor
+ clk_wzrd->clks_internal[wzrd_clk_mul] = clk_hw_register_fixed_factor
(&pdev->dev, clk_name,
__clk_get_name(clk_wzrd->clk_in1),
0, mult, div);
@@ -1092,10 +1096,8 @@ static int clk_wzrd_probe(struct platform_device *pdev)
}
clk_name = devm_kasprintf(&pdev->dev, GFP_KERNEL, "%s_mul_div", dev_name(&pdev->dev));
- if (!clk_name) {
- ret = -ENOMEM;
- goto err_rm_int_clk;
- }
+ if (!clk_name)
+ return -ENOMEM;
if (is_versal) {
edged = !!(readl(clk_wzrd->base + WZRD_CLK_CFG_REG(is_versal, 20)) &
@@ -1108,35 +1110,32 @@ static int clk_wzrd_probe(struct platform_device *pdev)
if (!div)
div = 1;
- clk_mul_name = __clk_get_name(clk_wzrd->clks_internal[wzrd_clk_mul]);
+ clk_mul_name = clk_hw_get_name(clk_wzrd->clks_internal[wzrd_clk_mul]);
clk_wzrd->clks_internal[wzrd_clk_mul_div] =
- clk_register_fixed_factor(&pdev->dev, clk_name,
- clk_mul_name, 0, 1, div);
+ clk_hw_register_fixed_factor(&pdev->dev, clk_name,
+ clk_mul_name, 0, 1, div);
} else {
ctrl_reg = clk_wzrd->base + WZRD_CLK_CFG_REG(is_versal, 0);
- clk_wzrd->clks_internal[wzrd_clk_mul_div] = clk_register_divider
+ clk_wzrd->clks_internal[wzrd_clk_mul_div] = clk_hw_register_divider
(&pdev->dev, clk_name,
- __clk_get_name(clk_wzrd->clks_internal[wzrd_clk_mul]),
+ clk_hw_get_name(clk_wzrd->clks_internal[wzrd_clk_mul]),
flags, ctrl_reg, 0, 8, CLK_DIVIDER_ONE_BASED |
CLK_DIVIDER_ALLOW_ZERO, &clkwzrd_lock);
}
if (IS_ERR(clk_wzrd->clks_internal[wzrd_clk_mul_div])) {
dev_err(&pdev->dev, "unable to register divider clock\n");
- ret = PTR_ERR(clk_wzrd->clks_internal[wzrd_clk_mul_div]);
- goto err_rm_int_clk;
+ return PTR_ERR(clk_wzrd->clks_internal[wzrd_clk_mul_div]);
}
/* register div per output */
for (i = nr_outputs - 1; i >= 0 ; i--) {
clkout_name = devm_kasprintf(&pdev->dev, GFP_KERNEL,
"%s_out%d", dev_name(&pdev->dev), i);
- if (!clkout_name) {
- ret = -ENOMEM;
- goto err_rm_int_clk;
- }
+ if (!clkout_name)
+ return -ENOMEM;
if (is_versal) {
- clk_wzrd->clkout[i] = clk_wzrd_ver_register_divider
+ clk_wzrd->clk_data->hws[i] = clk_wzrd_ver_register_divider
(&pdev->dev,
clkout_name, clk_name, 0,
clk_wzrd->base,
@@ -1148,7 +1147,7 @@ static int clk_wzrd_probe(struct platform_device *pdev)
DIV_O, &clkwzrd_lock);
} else {
if (!i)
- clk_wzrd->clkout[i] = clk_wzrd_register_divf
+ clk_wzrd->clk_data->hws[i] = clk_wzrd_register_divf
(&pdev->dev, clkout_name, clk_name, flags, clk_wzrd->base,
(WZRD_CLK_CFG_REG(is_versal, 2) + i * 12),
WZRD_CLKOUT_DIVIDE_SHIFT,
@@ -1156,7 +1155,7 @@ static int clk_wzrd_probe(struct platform_device *pdev)
CLK_DIVIDER_ONE_BASED | CLK_DIVIDER_ALLOW_ZERO,
DIV_O, &clkwzrd_lock);
else
- clk_wzrd->clkout[i] = clk_wzrd_register_divider
+ clk_wzrd->clk_data->hws[i] = clk_wzrd_register_divider
(&pdev->dev, clkout_name, clk_name, 0, clk_wzrd->base,
(WZRD_CLK_CFG_REG(is_versal, 2) + i * 12),
WZRD_CLKOUT_DIVIDE_SHIFT,
@@ -1164,22 +1163,20 @@ static int clk_wzrd_probe(struct platform_device *pdev)
CLK_DIVIDER_ONE_BASED | CLK_DIVIDER_ALLOW_ZERO,
DIV_O, &clkwzrd_lock);
}
- if (IS_ERR(clk_wzrd->clkout[i])) {
- int j;
-
- for (j = i + 1; j < nr_outputs; j++)
- clk_unregister(clk_wzrd->clkout[j]);
+ if (IS_ERR(clk_wzrd->clk_data->hws[i])) {
dev_err(&pdev->dev,
"unable to register divider clock\n");
- ret = PTR_ERR(clk_wzrd->clkout[i]);
- goto err_rm_int_clks;
+ return PTR_ERR(clk_wzrd->clk_data->hws[i]);
}
}
out:
- clk_wzrd->clk_data.clks = clk_wzrd->clkout;
- clk_wzrd->clk_data.clk_num = ARRAY_SIZE(clk_wzrd->clkout);
- of_clk_add_provider(np, of_clk_src_onecell_get, &clk_wzrd->clk_data);
+ clk_wzrd->clk_data->num = nr_outputs;
+ ret = devm_of_clk_add_hw_provider(&pdev->dev, of_clk_hw_onecell_get, clk_wzrd->clk_data);
+ if (ret) {
+ dev_err(&pdev->dev, "unable to register clock provider\n");
+ return ret;
+ }
if (clk_wzrd->speed_grade) {
clk_wzrd->nb.notifier_call = clk_wzrd_clk_notifier;
@@ -1198,25 +1195,6 @@ static int clk_wzrd_probe(struct platform_device *pdev)
}
return 0;
-
-err_rm_int_clks:
- clk_unregister(clk_wzrd->clks_internal[1]);
-err_rm_int_clk:
- clk_unregister(clk_wzrd->clks_internal[0]);
- return ret;
-}
-
-static void clk_wzrd_remove(struct platform_device *pdev)
-{
- int i;
- struct clk_wzrd *clk_wzrd = platform_get_drvdata(pdev);
-
- of_clk_del_provider(pdev->dev.of_node);
-
- for (i = 0; i < WZRD_NUM_OUTPUTS; i++)
- clk_unregister(clk_wzrd->clkout[i]);
- for (i = 0; i < wzrd_clk_int_max; i++)
- clk_unregister(clk_wzrd->clks_internal[i]);
}
static const struct of_device_id clk_wzrd_ids[] = {
@@ -1235,7 +1213,6 @@ static struct platform_driver clk_wzrd_driver = {
.pm = &clk_wzrd_dev_pm_ops,
},
.probe = clk_wzrd_probe,
- .remove_new = clk_wzrd_remove,
};
module_platform_driver(clk_wzrd_driver);
--
2.45.2General comment: do one thing in one patch, i.e. use clk_hw API and
don't also migrate to devm in one patch.
Quoting Harry Austen (2024-07-20 05:01:36)
> diff --git a/drivers/clk/xilinx/clk-xlnx-clock-wizard.c b/drivers/clk/xilinx/clk-xlnx-clock-wizard.c
> index 0ca045849ea3e..30c5cc9dcd7e9 100644
> --- a/drivers/clk/xilinx/clk-xlnx-clock-wizard.c
> +++ b/drivers/clk/xilinx/clk-xlnx-clock-wizard.c
> @@ -17,6 +17,7 @@
> #include <linux/of.h>
> #include <linux/math64.h>
> #include <linux/module.h>
> +#include <linux/overflow.h>
What is this include for? __counted_by()?
> #include <linux/err.h>
> #include <linux/iopoll.h>
>
> @@ -121,24 +122,22 @@ enum clk_wzrd_int_clks {
> /**
> * struct clk_wzrd - Clock wizard private data structure
> *
> - * @clk_data: Clock data
> + * @clk_data: Output clock data
> * @nb: Notifier block
> * @base: Memory base
> * @clk_in1: Handle to input clock 'clk_in1'
> * @axi_clk: Handle to input clock 's_axi_aclk'
> * @clks_internal: Internal clocks
> - * @clkout: Output clocks
> * @speed_grade: Speed grade of the device
> * @suspended: Flag indicating power state of the device
> */
> struct clk_wzrd {
> - struct clk_onecell_data clk_data;
> + struct clk_hw_onecell_data *clk_data;
It could be placed at the end and then one allocation could be used
instead of two.
> struct notifier_block nb;
> void __iomem *base;
> struct clk *clk_in1;
> struct clk *axi_clk;
> - struct clk *clks_internal[wzrd_clk_int_max];
> - struct clk *clkout[WZRD_NUM_OUTPUTS];
> + struct clk_hw *clks_internal[wzrd_clk_int_max];
> unsigned int speed_grade;
> bool suspended;
> };
> @@ -1108,35 +1110,32 @@ static int clk_wzrd_probe(struct platform_device *pdev)
> if (!div)
> div = 1;
>
> - clk_mul_name = __clk_get_name(clk_wzrd->clks_internal[wzrd_clk_mul]);
> + clk_mul_name = clk_hw_get_name(clk_wzrd->clks_internal[wzrd_clk_mul]);
> clk_wzrd->clks_internal[wzrd_clk_mul_div] =
> - clk_register_fixed_factor(&pdev->dev, clk_name,
> - clk_mul_name, 0, 1, div);
> + clk_hw_register_fixed_factor(&pdev->dev, clk_name,
> + clk_mul_name, 0, 1, div);
> } else {
> ctrl_reg = clk_wzrd->base + WZRD_CLK_CFG_REG(is_versal, 0);
> - clk_wzrd->clks_internal[wzrd_clk_mul_div] = clk_register_divider
> + clk_wzrd->clks_internal[wzrd_clk_mul_div] = clk_hw_register_divider
Are these going to be using devm so that on failure they get
unregistered?
> (&pdev->dev, clk_name,
> - __clk_get_name(clk_wzrd->clks_internal[wzrd_clk_mul]),
> + clk_hw_get_name(clk_wzrd->clks_internal[wzrd_clk_mul]),
> flags, ctrl_reg, 0, 8, CLK_DIVIDER_ONE_BASED |
> CLK_DIVIDER_ALLOW_ZERO, &clkwzrd_lock);
> }
> if (IS_ERR(clk_wzrd->clks_internal[wzrd_clk_mul_div])) {
> dev_err(&pdev->dev, "unable to register divider clock\n");
> - ret = PTR_ERR(clk_wzrd->clks_internal[wzrd_clk_mul_div]);
> - goto err_rm_int_clk;
> + return PTR_ERR(clk_wzrd->clks_internal[wzrd_clk_mul_div]);
> }
On Wed Jul 24, 2024 at 12:14 AM BST, Stephen Boyd wrote:
> General comment: do one thing in one patch, i.e. use clk_hw API and
> don't also migrate to devm in one patch.
Fair point. Will split into two patches in v2.
>
> Quoting Harry Austen (2024-07-20 05:01:36)
> > diff --git a/drivers/clk/xilinx/clk-xlnx-clock-wizard.c b/drivers/clk/xilinx/clk-xlnx-clock-wizard.c
> > index 0ca045849ea3e..30c5cc9dcd7e9 100644
> > --- a/drivers/clk/xilinx/clk-xlnx-clock-wizard.c
> > +++ b/drivers/clk/xilinx/clk-xlnx-clock-wizard.c
> > @@ -17,6 +17,7 @@
> > #include <linux/of.h>
> > #include <linux/math64.h>
> > #include <linux/module.h>
> > +#include <linux/overflow.h>
>
> What is this include for? __counted_by()?
For struct_size()
>
> > #include <linux/err.h>
> > #include <linux/iopoll.h>
> >
> > @@ -121,24 +122,22 @@ enum clk_wzrd_int_clks {
> > /**
> > * struct clk_wzrd - Clock wizard private data structure
> > *
> > - * @clk_data: Clock data
> > + * @clk_data: Output clock data
> > * @nb: Notifier block
> > * @base: Memory base
> > * @clk_in1: Handle to input clock 'clk_in1'
> > * @axi_clk: Handle to input clock 's_axi_aclk'
> > * @clks_internal: Internal clocks
> > - * @clkout: Output clocks
> > * @speed_grade: Speed grade of the device
> > * @suspended: Flag indicating power state of the device
> > */
> > struct clk_wzrd {
> > - struct clk_onecell_data clk_data;
> > + struct clk_hw_onecell_data *clk_data;
>
> It could be placed at the end and then one allocation could be used
> instead of two.
Ah good point. Will move in v2.
>
> > struct notifier_block nb;
> > void __iomem *base;
> > struct clk *clk_in1;
> > struct clk *axi_clk;
> > - struct clk *clks_internal[wzrd_clk_int_max];
> > - struct clk *clkout[WZRD_NUM_OUTPUTS];
> > + struct clk_hw *clks_internal[wzrd_clk_int_max];
> > unsigned int speed_grade;
> > bool suspended;
> > };
> > @@ -1108,35 +1110,32 @@ static int clk_wzrd_probe(struct platform_device *pdev)
> > if (!div)
> > div = 1;
> >
> > - clk_mul_name = __clk_get_name(clk_wzrd->clks_internal[wzrd_clk_mul]);
> > + clk_mul_name = clk_hw_get_name(clk_wzrd->clks_internal[wzrd_clk_mul]);
> > clk_wzrd->clks_internal[wzrd_clk_mul_div] =
> > - clk_register_fixed_factor(&pdev->dev, clk_name,
> > - clk_mul_name, 0, 1, div);
> > + clk_hw_register_fixed_factor(&pdev->dev, clk_name,
> > + clk_mul_name, 0, 1, div);
> > } else {
> > ctrl_reg = clk_wzrd->base + WZRD_CLK_CFG_REG(is_versal, 0);
> > - clk_wzrd->clks_internal[wzrd_clk_mul_div] = clk_register_divider
> > + clk_wzrd->clks_internal[wzrd_clk_mul_div] = clk_hw_register_divider
>
> Are these going to be using devm so that on failure they get
> unregistered?
I appear to have missed that entirely. Yes, that was the intention. Will fix in v2.
>
> > (&pdev->dev, clk_name,
> > - __clk_get_name(clk_wzrd->clks_internal[wzrd_clk_mul]),
> > + clk_hw_get_name(clk_wzrd->clks_internal[wzrd_clk_mul]),
> > flags, ctrl_reg, 0, 8, CLK_DIVIDER_ONE_BASED |
> > CLK_DIVIDER_ALLOW_ZERO, &clkwzrd_lock);
> > }
> > if (IS_ERR(clk_wzrd->clks_internal[wzrd_clk_mul_div])) {
> > dev_err(&pdev->dev, "unable to register divider clock\n");
> > - ret = PTR_ERR(clk_wzrd->clks_internal[wzrd_clk_mul_div]);
> > - goto err_rm_int_clk;
> > + return PTR_ERR(clk_wzrd->clks_internal[wzrd_clk_mul_div]);
> > }
© 2016 - 2025 Red Hat, Inc.